Mapping the 5th CD primary

No, I'm not back -- not really. But I did want to say a few words about the 5th CD race.

You see, the Strib put together some nifty-looking maps, along with an article explaining them. And they sort of mucked it up. So I figured that since I'm the map guy, I ought to set the record straight.

The Strib maps look awesome. The color choices are good, they show the city boundaries, there's a little inset map that shows where the district is in relation to the metro area. And they even superimposed Ellison's, Ostrow's, and Reichgott Junge's districts as an added bonus. Great work on the presentation.

The problem is that they mapped the wrong thing. They mapped the percentage of the vote in each precinct. This led their reporter, looking at the maps, to make some statements that just don't fit the facts very well.

The 'hook' of the story was that Ellison's victory was due to the Minneapolis vote -- that he lost big in the suburbs:

Keith Ellison's impressive primary win in the Fifth Congressional District was made possible by strong support across Minneapolis -- including its poorest and most affluent neighborhoods -- that offset his weaker showing in the suburbs.

A closer look at Tuesday's DFL primary results showed that the north Minneapolis state legislator not only won in areas of the city where he was expected to do well, but also scored a notable victory in northeast Minneapolis, a presumed stronghold of candidate Paul Ostrow, a Minneapolis City Council member.

One of seven candidates, Ellison won a majority of the vote in north, south and southeast Minneapolis. In south Minneapolis, where the highest percentage of registered voters in the district went to the polls, Ellison won 56 percent of the vote. In southwest Minneapolis, which has many of the city's most affluent neighborhoods, Ellison finished first with 45 percent of the vote.

In north Minneapolis, the location of much of the city's minority population, Ellison walked away with 66 percent of the vote.

But in suburbs, stretching from Richfield to Spring Lake Park, Ellison captured no more than 26 percent of the total. The district's northern suburbs were a particular struggle. He garnered 17 percent of the vote in Columbia Heights, and just 14 percent in New Hope. His strongest suburban showing was in Hopkins, where he still finished third behind Mike Erlandson and Ember Reichgott Junge.

Everything here is true, but it glides by the most important fact -- that turnout in South and Southwest Minneapolis was much higher than anywhere else. Here's a turnout map (with a population map for comparison):
 

 
It's true that voters in Anoka county didn't vote for Keith Ellison. But they didn't vote against him, either -- they basically stayed home. The same was true for voters throughout the suburbs, with the notable exceptions of Golden Valley, Richfield, and the northeast corner of St. Louis Park. Ellison's strong showing in the city didn't "offset" high vote totals for his opponents in the suburbs, because no one got particularly high vote totals in the suburbs.

Ellison and Erlandson did the best job of turning out voters, and the voters they both turned out were in South and Southwest Minneapolis. And practically everywhere in the city that Erlandson turned out a lot of voters, Ellison's campaign turned out more. Maps of the raw vote are quite different than the percentage maps that the Strib printed:
 

 
Looking at these maps, the Strib analysis is at best incomplete. Erlandson and Reichgott Junge did outpoll Ellison in the suburbs, but their strongest support was limited to a few pockets here and there. If Erlandson had truly run as well in the suburbs as the percentage maps seem to suggest, with those big dark blotches in Anoka County and St. Louis Park, the election would have been much closer.

It's not just about drawing good maps -- it's also about choosing which maps to draw.

posted by DK @ 2:59 AM permalink